Also if you don't know what your target palette is you can do k-means clustering over colorspace, and the palette is the cluster centers (each point you're giving to k-means is an <r,g,b> or <h,s,v> vector).

K-medians or k-medoids yield more pleasing results in my experience

For small distances, yes. If you plot a 2d projection of a dataset that doesn't have much structure you're going to be reading patterns into whitenoise (though this data has some pretty clear clusters, which are probably real). If I were doing something other than writing a fun blog post I would have done cluster analysis with something like DBSCAN.

There's no UI for if there are no matches; it just does nothing. Try searching for \.com or something.

Edit: I patched it so it displays an alert if there are no matches.

I see. That patch makes it a lot nicer to find out that none of the sites i wanted to look for show up in the data :)

Non-spam bots generally don't follow each other or link to external websites. (I'm also the author of one of the more popular image bots https://twitter.com/a_quilt_bot)

(original author here) it's a spambot network. If you click the link in that post to the interactive version (this: https://pile-of-junk.s3.amazonaws.com/twitter_scatter_10k.ht...) you can see for yourself.

People in this position generally don't start successful companies. Do charity work with the money instead.

Dogecoin is a good, not a medium of exchange. Why do people buy iPhone apps? iPhone apps don't even have resale value.

The word you're looking for is "asset", and mediums of exchange are by definition always assets. When people talk about their aspirations for Bitcoin, the most techno-utopians want it to become a commonly accepted store of value or unit of account.

Keep in mind that mediums of exchange can be seen as purely "any mechanism that can signal the irrevocable transfer of the value". The innovation btc brings to this is instead of transferring physical items you create cryptographic hashes that you then publicize.

It's worth noting that when we talk about money we're trying to describe emergent phenomena. This is what we've got, and the following items are just useful ways of thinking about the issue - except unlike in physics we don't have the leeway to conduct careful, repeatable experiments.

http://en.wikipedia.org/wiki/Asset http://en.wikipedia.org/wiki/Medium_of_exchange http://en.wikipedia.org/wiki/Store_of_value http://en.wikipedia.org/wiki/Unit_of_account

I forget the (accounting/financial) term for talking expenditures like videogames or movies or haircuts. I'm pretty sure we can describe those all as "services" but I've yet to find the appropriate wikipedia article. If someone can jog my memory that'd be great.

How is Dogecoin a good if its only feature is the mode by which it is exchanged?

Disclaimer: I possess Dogecoins

It makes the people that hold it laugh, which is valuable to them.

except they do have a resale value.

Try reselling Angry Birds.

sure - apps dont have a resale value. but that doesnt mean video games in general dont. You're just providing an exception to a rule that's all. Also - you're defying your original point of something being a good. Angry birds is not a good - its a service.

We set the maximum seal-time to be relatively short for practical reasons on our end (we have to keep the capsules alive), but coming up with ways to extend that time is something we're thinking about.

This is why I get annoyed with the current trend against doing technically hard things as startups. The up-front risk of your thing not working is much higher, but if it does (and there's a market) then you're in a much more defensible position. You're basically front-loading risk. And since the thing you're risking is time, front-loading is something you should do if you can.

There is the problem of telling whether there's a market for a given thing before you invest six months or a year building it, so I'm not advocating sinking a whole year without shipping, but there's definitely merit to starting with something that can't be built as a rails app.

All of those products exist because they all have slightly different benefits and drawbacks. What cocktail of analytics products is right for you depends on your specific needs. In general I would say use google analytics (because it's free), and when you want to pull a report that google analytics can't give you, consider what other service(s) could.

Completely agree, which is why I asked the question. I wanted to hear more about the analytics cocktails of others and the startup they are working on to determine which ones may be useful to investigate further if their startup is analogous in someway to the one I'm working at now.